Crazy Transit Pitches

Tunneling in the water is so expensive that it's too crazy for crazy.
There are 5 subway tunnels crossing the East River in NYC. There is plenty of precedent.

I wasn't talking about tunneling though. A tie-in to the Grand Junction/BU bridge underpass above ground is much more feasible.
Not proposing my own urban ring route here. Just a reaction to that map.
 
All four of Boston's rail transit systems cross at least one body of water. Orange crosses the Charles and Mystic rivers; Red crosses the Charles and Neponset rivers; Green crosses the Charles river; Blue crosses the harbor. That's six water crossings, arguably five of them in a nort/south dimension. I'm not sure what you are looking for in terms of more.
 
There are 5 subway tunnels crossing the East River in NYC. There is plenty of precedent.

I wasn't talking about tunneling though. A tie-in to the Grand Junction/BU bridge underpass above ground is much more feasible.
Not proposing my own urban ring route here. Just a reaction to that map.

There are a lot of good GLX-UR proposals for the Grand Junction that are *way* better than the hackneyed Commuter Rail official proposals.

As far as river crossings, as noted there are many, but there aren't any of the Charles River west of Charles Circle, which I think is what you're noting. A lot of that is because tunneling outside of the urban core wasn't on the radar in the early-20th Century due to the prevalence of streetcars. No Mass Ave subway, no subway between MIT and Back Bay, etc. At this point the ship has really sailed because of the challenges of tunneling through Boston's glacial till geography. Too risky to the architecture to do a lot of cut-cover in the landfill zones, and too risky/costly to bore through/beneath the upper layers of earth. New tunneling is easiest in places where cut-cover is practical or where there's relatively shallow bedrock to take advantage of with a TBM.
 
All four of Boston's rail transit systems cross at least one body of water. Orange crosses the Charles and Mystic rivers; Red crosses the Charles and Neponset rivers; Green crosses the Charles river; Blue crosses the harbor. That's six water crossings, arguably five of them in a nort/south dimension. I'm not sure what you are looking for in terms of more.
Those are still all in effect of Hub+Spoke. The East-West focus was in regards to the zoom-in of the Brookline/Back Bay/South Boston area done by Riverside. Super inefficient routing going inbound then outbound in many instances.

Yeah, plenty of people here have found great GLX options to solve this. I understand the crazy logistical issues of tunneling here.
I think it's been discussed before (maybe?) but is an LRT line along Mass Ave prohibitive? Something about the proximity of existing intersections and backup.
 
Tunneling in the water is so expensive that it's too crazy for crazy.
Pardon my ignorance, but is this true? My understanding is that the Ted Williams Tunnel was one of the relatively least expensive, greatest bang-for-the-buck components of the Big Dig (Ari Ofsevit alludes to this in his pitch for a BL tunnel extension parallel to the Longfellow from Charles/MGH to the Volpe site at Kendall). Sure, tunneling under water isn't cheap, but neither is tunneling under a centuries-old city. How does the cost of sinking pre-built tubes in a trench in the (polluted but otherwise free of obstructions) Charles River differ from using a TBM or even doing cut-and-cover across densely developed urban land with centuries-old mystery utilities underneath?
There are a lot of good GLX-UR proposals for the Grand Junction that are *way* better than the hackneyed Commuter Rail official proposals.
Yeah, running the GL across the Grand Junction using the existing rail bridge under the BU bridge is a very doable project (see F-Line's post here for some great sketches). You could connect Lechemere + East Cambridge + Kendall + MIT + Cambridgeport to BU + Allston/Brighton + future West Station using mostly existing infrastructure.

Another real possibility for a river crossing is connecting the Cambridge and Boston sides of Harvard's campus (again, F-Line's post here has some great sketches) through a tunnel parallel to the Anderson Memorial Bridge around Harvard Stadium. This could also use some of the existing but currently-unused RL tunnel from before the 1980s Alewife extension to connect to the exiting Harvard Station on the Cambridge side. That'd give Cambridge Red Line riders (plus suburban commuters from Alewife) a much more direct connection to Allston/Brighton, the Fenway, and even parts of the Back Bay without having to go through the mess at Park St.
 
Would a Green Line branch to Allston make more sense? Just continue onto Brighton Ave from Commonwealth B line. A separate Blue extension that way seems redundant.
Blue running down Brookline Ave would be nice to give actual Fenway the Fenway stop (it's currently weirdly out of the way and not central to all the density being built down Brookline and Boylston).

Is there a reason for doubling up lines on top of each other other than expediency of the below grade one? Or are you suggesting just replacing the Green E entirely with Blue subway?

Replying here. "Blue-to-Allston" isn't so much about bringing the Blue Line to Allston specifically but for opening it up to continue on to places like Harvard, Newton Corner, Watertown, Waltham, and Riverside. Green Line branch(es) itself are almost certainly the better play for service to Allston itself, but if you wanted to go farther than Newtown Corner, I think it becomes less appealing, which is why it would be a question of whether Blue is better (especially if Blue runs largely express through Allston via a Mass Pike alignment). Hardly a sure thing, but that's why it's typically separated from the more modest Blue-to-Kenmore extension, especially since...

...Blue extensions to the south of Kenmore are also a mixed bag. Believe me, you don't need to sell me on the raw benefits of a subway down Brookline Ave -- the tragedy of transit access to LMA is a repeated saga of "close but not quite there". But I can't even see a subway being built there, if for no other reason than you have a perfectly good ROW just to the west. Yeah, the walking catchments overlap poorly because of the park in between, but I can't imagine the benefits outweighing the costs.

I think it'd be a better choice to put bus lanes and related infrastructure down Brookline Ave and Longwood Ave and do our best to make that be BRT-esque.

Blue to Brookline Village via the existing LRT ROW isn't a bad idea, but it has its own problems, chief among them being that you can't go any further out on the Highland Branch: any future extension to Needham is going to involve several grade crossings in Downtown Needham, and those will require LRT rather than HRT. If you run Blue past Brookline Village, you either block or hamstring severely an extension to Needham (which is badly needed in order to free up space on the Northeast Corridor).

I don't think I was clear earlier -- definitely not proposing replacing the Green E with the Blue Subway. Rather, I'm proposing that Surface Green E be replaced with an extended Subway Green E + D: extend the Huntington Subway down to Riverway, and then connect it to the D Line at Brookline Village, and extend the E to Needham.

In general, I agree, it's better to avoid doubling up lines. However, it's worth saying that the Green Line makes for a theoretical exception to that rule in some cases, because a surface level Green Line could act as a "local" service while a subway underneath could run as "express." The Green Line is always a little bit of an exception to every rule because it can exist on a sliding scale where you have HRT-style rapid transit on one end and local bus on the other.

Case in point -- if the Blue did get extended to Allston, I would imagine that it doesn't replace any particular Green Line branch, but rather operates as an express bypass.
 
BLX-Allston would get anywhere it goes by banging a right from Brookline Ave. under the B&A. Including if it swallowed the D...it would take the B&A jog to insert without severing the C tunnel. So kind of by necessity it's a separate deal from GL. You're avoiding Beacon Jct. to save the C so any intermediate stops @ Landsdowne CR would be BLX-only; Beacon St. Eversource substation blocks a C/D interface. Next crossing is BU Bridge Jct. where the B St. Paul portal tunnel is actively inclining under the Pike (and BLX is slipping under that)...no opportunity for a transfer station.

So West would be the first common station opportunity after Kenmore...and there you're trapping Kendall-routed UR patterns for a different audience mix. Ring thru-routes from Kendall to Harvard are part/parcel the value proposition of the UR filets, so you aren't mode-changing the rest of the Harvard Branch to HRT. You're rather continuing somewhere west from West along the B&A to cover a different swath of Allston. And by that point you're on totally different/unique corridor from the B, A/57, and Harvard Branch.

So it reasons to assume that any west-of-Kenmore TBD's are in addition to Green/LRT, not swallowing any. Which should not be surprising, as the 1945 BTC expansion plan traced out HRT to Riverside via Newton Corner as wholly additional to the as-planned LRT D Line. Obviously now with Pike eating ex-B&A Tks. 3 & 4 it's now a subwaying prospect instead of surface grade separated, but same (or similar w/ different outer terminus) build is straightforwardly feasible today given appropriate resources and ROI.
 
BLX-Allston would get anywhere it goes by banging a right from Brookline Ave. under the B&A. Including if it swallowed the D...it would take the B&A jog to insert without severing the C tunnel. So kind of by necessity it's a separate deal from GL. You're avoiding Beacon Jct. to save the C so any intermediate stops @ Landsdowne CR would be BLX-only; Beacon St. Eversource substation blocks a C/D interface. Next crossing is BU Bridge Jct. where the B St. Paul portal tunnel is actively inclining under the Pike (and BLX is slipping under that)...no opportunity for a transfer station.

So West would be the first common station opportunity after Kenmore...and there you're trapping Kendall-routed UR patterns for a different audience mix. Ring thru-routes from Kendall to Harvard are part/parcel the value proposition of the UR filets, so you aren't mode-changing the rest of the Harvard Branch to HRT. You're rather continuing somewhere west from West along the B&A to cover a different swath of Allston. And by that point you're on totally different/unique corridor from the B, A/57, and Harvard Branch.

So it reasons to assume that any west-of-Kenmore TBD's are in addition to Green/LRT, not swallowing any. Which should not be surprising, as the 1945 BTC expansion plan traced out HRT to Riverside via Newton Corner as wholly additional to the as-planned LRT D Line. Obviously now with Pike eating ex-B&A Tks. 3 & 4 it's now a subwaying prospect instead of surface grade separated, but same (or similar w/ different outer terminus) build is straightforwardly feasible today given appropriate resources and ROI.

Would it ever make sense to branch the Blue Line after Kenmore? Branch 1 some Allston-to-who-knows-where trajectory, as you describe, and Branch 2 continuing down Brookline Ave to terminate at Brookline Village, retiring D along the Muddy River and pushing D/Needham to a Huntington subway? I recall previously you made it clear that branching any HRT line must occur after the line's major bus terminal. While not on the same scale as Malden or Forest Hills, Kenmore seems to be that bus terminal for an extended BL (or maybe it would be West Station in the future?). I feel that LMA is too important regionally, has too many jobs, and serves a population (i.e. patients) where the current walkshed to D is too much. I know the MBTA just finished its flood protection project at the D portal, but taking rapid transit out of that corridor and retiring that portal seems like a win for resiliency.
 
Yes...if Kenmore is the last trunk transfer well of heft branching would be acceptable.

What you'd then have to study is that both Branch 1 and Branch 2 agree with each other on that being the last gravity well that swings trunkline heft. If Branch 1 runs to Newton Corner then turns under Galen St. to Watertown, however, the West region-largest bus hub of Corner+H2O might not agree with frequency halving. That's just one example where too-early branching could *potentially* backfire, so make sure your 'vision thing' is clear about where both branches are going and what that accomplishes before biting the bullet.
 
Probably not LRT ever. Even true BRT is tricky given the volumes of traffic that Mass Ave receives in certain sections.

Would it be crazy expensive to tunnel under Mass Ave and the river? Be cool to have a direct, traffic-avoiding connection between Harvard and Nubian. It’d intersect so many of the East-West lines.
 
Last edited:
I think the best that can be hoped for on the 1 is BRT elements without full BRT. But I'm a little less pessimistic about it than I might be. It would probably take another 10 years of generational and political change, but the road is very wide for virtually the entire stretch. If you can build a political consensus to eliminate parking along certain stretches, you could probably swing it.
 
Would it be crazy expensive to tunnel under Mass Ave and the river? Be cool to have a direct, traffic-avoiding connection between Harvard and Nubian. It’d intersect so many of the East-West lines.

It's a tough subterranean geography for tunneling because of the lack of bedrock anywhere near the surface for boring. Cut-cover has all the typical utility and building foundation challenges, but not impossible. So it's not going to be on anyone's realistic Top 10 expansion projects. The biggest issue for me is where is it coming from? Red claims Mass Ave until it veers off towards Kendall and a southward Mass Ave subway isn't going to be able to split off of that. So where does this subway go north of MIT?
 
So where does this subway go north of MIT?

Does a subway line need to be long in order to be useful? Berlin's U4 line only has 5 stations.

Even with just Nubian, Mass Ave (OL), Hynes (GL), maybe a BLX station near the river, and Central (RL) that's the same number of stations and it links practically all the lines without going downtown. Maybe add a stop at BMC and/or Vassar Street for good measure.

If you do want it to go somewhere, though, maybe have it go south of Nubian (under Dudley Street) to link up with the Fairmount Line at Uphams?

Or did you mean something more like "where do the trains turn around and stay when not in use"? In which case, I'm not sure. I don't know enough about the field. I'd be interested to find out how my small line example turns around at Nollendorfplatz and Innsbrucker Platz if a similar thing at Central and Nubian would be too difficult.
 
Last edited:
Does a subway line need to be long in order to be useful? Berlin's U4 line only has 5 stations.

Even with just Nubian, Mass Ave (OL), Hynes (GL), maybe a BLX station near the river, and Central (RL) that's the same number of stations and it links practically all the lines without going downtown. Maybe add a stop at BMC and/or Vassar Street for good measure.

If you do want it to go somewhere, though, maybe have it go south of Nubian (under Dudley Street) to link up with the Fairmount Line at Uphams?
Kendall would be better than Central. Tack on Lechmere and Sullivan for extra credit. But that idea works well as BRT-ish. It certainly doesn’t need to be HRT. There have been lots of ideas in this and other threads over the years including branches off of the Mass Ave BRT spine (e.g. to Kenmore and Brookline Ave)
 
It's a tough subterranean geography for tunneling because of the lack of bedrock anywhere near the surface for boring. Cut-cover has all the typical utility and building foundation challenges, but not impossible. So it's not going to be on anyone's realistic Top 10 expansion projects. The biggest issue for me is where is it coming from? Red claims Mass Ave until it veers off towards Kendall and a southward Mass Ave subway isn't going to be able to split off of that. So where does this subway go north of MIT?

Pre-1905 the Back Bay shoreline on the Cambridge side was Main & Sidney Streets, so the Red Line already traced out the limits of the ancestral shoreline bedrock on its route to the Longfellow. Every single block of MIT to the south of there has only existed for 115 years as very soft fill, with the Grand Junction RR being the only older extant structure as it was originally laid out in the middle of the Bay on an earthen causeway with multiple bridges crossing over marshy spits. The subsurface fill out there is very low-density, so for 2/3 mile underneath Mass Ave. to Memorial Drive you're not only going to have manifold difficulties waterproofing a subway tunnel...but also will have problems anchoring it from slight lateral movement. It'll almost have to be designed like a cross-Harbor bore a la the Blue Line where the sunk sections have built in flex under slight movement to the whole tube buried inside the silt. And it'll be way harder to bury flex sections under built-up linear street than it is to work underneath open Harbor, so construction staging is going to be a blowout unto itself. Then of course you have to somehow insert stations in that thing while dealing with those build properties. And calculate how far on the Boston side you have to continue the special mitigation under the somewhat earlier Back Bay neighborhood fill.

This is the same reason why every Crazy "why don't we just bury the Grand Junction?" Pitch is a tripleplus NO! on feasibility. Same porous shit because the land around the original Boston & Albany causeway was just backfilled around the active line, only now you're also baking sharp curves and a dangerous 'storm drain' underpin of Red @ Kendall into the difficulty mix.

It's not unreasonable to speculate that a Mass Ave. subway, no matter how achingly perfect it looks on a map, could indeed cost just as much as NSRL. And certainly more than any other individual transit project ever. While in no way/shape/form doing as much for us as the very scarily expensive NSRL or other very very expensive most-wanteds. Unfortunately that is realm-of-impossibility stuff on cost/benefit. Although, keep in mind that a streamlined 1/CT1 with Urban Ring LRT picking up @ Mass Ave. is cosmically better than today. If we simply densify the number of rapid transit touches crossing Mass Ave.--UR @ Harvard, UR @ MIT, possibly BLX-Kenmore's Beacon St. intermediate in the Storrow cut, more varieties of GL service patterns run thru @ Hynes & Symphony from the various interconnects, OL Mass Ave. with 3 min. headways & outer-neighborhoods extension, Nubian streetcar touch, and Newmarket Urban Rail + UR Southeast quadrant BRT spur to JFK--you keep the 1/CT1 thoroughly and constantly overchurning its ridership. Very few people will need to be riding it end-to-end compared with today because of all the radial augments. So the frequent transfer touches is arguably the right feasible move for us to be pursuing now, because it'll keep us from having to consider this a full-on linear corridor where obviously the build options are physically very limited and unlikely to ever be feasible. Rather than flailing at the unanswerable build...these new radials simply change the set of prelim questions about what kind of corridor the 1 exactly is. The way it shapes up, that looks to be plenty good enough for the task if done in conjunction with some beneficial bus-ops streamlining.
 
Does a subway line need to be long in order to be useful? Berlin's U4 line only has 5 stations.

Even with just Nubian, Mass Ave (OL), Hynes (GL), maybe a BLX station near the river, and Central (RL) that's the same number of stations and it links practically all the lines without going downtown. Maybe add a stop at BMC and/or Vassar Street for good measure.

If you do want it to go somewhere, though, maybe have it go south of Nubian (under Dudley Street) to link up with the Fairmount Line at Uphams?

Or did you mean something more like "where do the trains turn around and stay when not in use"? In which case, I'm not sure. I don't know enough about the field. I'd be interested to find out how my small line example turns around at Nollendorfplatz and Innsbrucker Platz if a similar thing at Central and Nubian would be too difficult.

I think the problem of where such a line would terminate on the northern end is indicative of another reason why the corridor isn't right for HRT, and would be debatable for any construction project that involves significant digging.

From Central Square, there are basically three directions to go: to the right, toward Inman and Union Squares; straight ahead toward Harvard; and to the left, toward Allston and Watertown.

Allston/Watertown would combine circumferential Mass Ave service with radial service to the western suburbs, which isn't a great combination -- you'd basically be marrying together two different kinds of lines, just because they happen to meet up. Harvard is a tough sell because it duplicates the Red Line, and even once you get to Harvard, you run into the same radial vs circumferential mismatch in terms of where you go further.

Inman/Union would be a plausible candidate, at least from a mapping perspective... but it would be a pretty serious tunneling project, and would significantly duplicate LRT service on the Grand Junction (much easier to build), in terms of integration to the larger network. (Especially since the logical destination after Union would be Sullivan.)

And, not for nothing, but the corridor traversed by the 1 isn't "Central-to-Nubian" but rather "Harvard-to-Nubian". Check out the Better Bus Profile. Something like 940 riders board at/around Harvard, followed by perhaps another 400 who board north of Central. It makes sense that there would be a large pool of Harvard boardings -- a lot of them are probably transfers from one of the many buses that terminate there. So a subway line that terminates at Central would still be missing a fair number of riders.
 
I think the problem of where such a line would terminate on the northern end is indicative of another reason why the corridor isn't right for HRT, and would be debatable for any construction project that involves significant digging.

From Central Square, there are basically three directions to go: to the right, toward Inman and Union Squares; straight ahead toward Harvard; and to the left, toward Allston and Watertown.

I’m still confused by why it would need to go somewhere else rather than just turn around at Central or Harvard? Also, why would it need to be HRT vs buried LRT?
 
Given that we're in "crazy" transit pitches, might I pitch that if we're tunneling for this line, we may as well tunnel to Back Bay Station, as opposed to Mass Ave Station. Span the platforms of the southern terminus, beneath the Mass Pike, between Copley and Back Bay, adding a subterranean, post-fare-gate transfer between IB Green, OB Green, OL, and the new station. That connectivity to higher ridership stations (plus better connectivity with Commuter Rail and Amtrak), is worth considering if you're already doing a crazy-build tunnel. It doesn't need to follow the above-ground Mass Ave routing.
 

Back
Top